A Chinese fuzzy autocompletion approach

Autocompeltion is a feature for predicting a word or phrase that the user wants to type in without actually typing it completely. Fuzzy autocompletion technologies have been presented recently to deal with the typing error. However, they do not perform well in Chinese circumstance. In the paper, it extends autocompletion by (1) supporting Chinese character fuzzy match (2) considering phonetic similarity and shape similarity of Chinese characters(3) allowing infix match as well as prefix match. When typing a Chinese character string, minor mistakes are tolerated and expected autocompletion strings are presented to user. Our empirical evaltuation demonstrates that our autocompletion algorithm has high-real-time performance to meet the demand of user interaction. We have developed a autocompletion framework using this algorithm and has been used in a Traditional Chinese medicine (TCM) application.